volume 8 Number 21 1980 Nucleic Acids Research

Purification and properties of DNA endonudease associated with Friend leukemia virus

Jon Nissen-Meyer and Ingolf F.Nes*

Department of Biochemistry, University of Bergen, Arstadveien 19, N-5000 Bergen, Norway

Received 14 July 1980

ABSTRACT

An endonudease associated wi th the core of Friend leukemia v i rus(FLV) has been pur i f ied more than 103- fo ld by ion exchange chromatographyand gel f i l t r a t i on . Its molecular weight was determined by gel f i l t ra t ion to beabout 40,000. Divalent cations were required for the endonudease to funct ionand KCI concentrations above 50 mM inhib i ted the enzyme ac t iv i t y . In thepresence of Mg the pur i f ied enzyme nicked preferent ia l ly supercoiledc i rcu lar DNA duplexes and in most of these molecules only one s ingle-stranded nick was introduced per s t rand . The regions into which the nickcould be i n t r oduce^ appeared to be randomly, d is t r ibu ted on the c i rcu larmolecule. When Mn was subst i tu ted for Mg the number of nicks i n t r o -duced into DNA by the pur i f ied enzyme was great ly increased, and bothrelaxed c i rcu lar and linear DNA duplexes were nicked as well as supercoiledc i rcu lar DNA duplexes. Prior to its pur i f i ca t ion , however, in the presence ofMn the endonudease act iv i ty in the v i rus ext ract was able to d i f ferent ia tebetween c i rcu lar and linear DNA duplexes, since both supercoiled andrelaxed c i rcu lar duplexes were nicked much more readi ly than linear dup lex-es. Single-stranded DNA funct ioned poorly as a substrate for the pur i f iedenzyme.

INTRODUCTION

The importance of a v i rus endonudease in the replication of re t ro -

v i ruses is suggested by the fact that such an enzyme is present in the v i ra l

core of the avian myeloblastosis v i rus (AMV) ( 1 ) . I t appears that th is ac t i -

v i t y resides in both the p-subunit of the v i ra l RNA-directed DNA polymerase

(2 ,3 ) and in the p32 pro te in , which is der ived by proteolyt ic cleavage of the

polymerase p-subuni t ( 1 ) . In the presence of divalent cations the endo-

nudease act iv i ty associated wi th the AMV p32 protein converts supercoiled

c i rcu lar DNA duplexes (RFI ) to the relaxed form (RF I I ) by in t roduc ing

single-st randed nicks into the DNA ( 1 ) . When Mg4"1" is present, the enzyme

introduces in most of the supercoiled duplexes only one nick per s t r and ,

whereas a large number of nicks are int roduced into both strands of the

supercoiled DNA in the presence of Mn ( 1 ) . In contrast to c i rcu lar dup lex -

© IRL Press Limited. 1 Falconberg Court. London W1V 5FG, U.K. 5043

Downloaded from https://academic.oup.com/nar/article-abstract/8/21/5043/1063672by gueston 09 April 2018

Nucleic Acids Research

es, linear duplexes (RFIII) and single-stranded DNA function relatively

poorly as substrate for the enzyme (1).

We recently reported the presence of an endonuclease activity in crude

extracts of Friend leukemia virus (FLV) (4). The activity in the FLV extract

was shown to have all of the above properties which are characteristic of the

endonuclease activity of the AMV p32 protein. The fact that two retroviruses

from different species appear to contain very similar endonucleases suggests

that they are bonafide virus enzymes which are important in the biology of

these viruses. This notion is also supported by the fact that the AMV p32

protein is coded for by the AMV genome. In the present work we report the

purification of the FLV associated endonuclease and the characterization of

the purified enzyme.

MATERIALS AND METHODS

Cell line, culture conditions and virus purification. FLV was propagated

in Eveline cells (5). The cells were maintained in suspension culture in

Eagle's minimum essential medium with Earle's balanced salt solution and 10%

foetal calf serum (Gibco Biocult). The medium was supplemented with 100

units/ml of penicillin and 100 ug/ml of streptomycin. The cell density was

kept at 2 to 4 x 105 cells/ml by adding an equal volume of fresh medium

every 24 h. The FLV was purified from the cell culture as earlier described

(4) .

Purification of virus associated endonuclease. Purified viruses at a

concentration of 1 mg protein/ml in buffer A (20 mM Tris-HCI (pH 7.5),

0.1 mM EDTA and 10 mM p-mercaptoethanol) were disrupted by adding Triton

X-100 to a final concentration of 0.2%. Two volumes of buffer A were added

to the virus extract after incubating the extract for 30 min at 0°C. The

extract was then applied to a DEAE-cellulose column (about 1 mg of protein

per ml bed volume) which was subsequently washed with two volumes of

buffer A. The proteins were eluted from the column with a linear gradient

(10 column volumes) of 0-600 mM NaCI in buffer A, and fractions equivalent

to 4% of the gradient volume were collected and assayed for endonuclease

activity. The endonuclease which eluted at about 50 mM NaCI was then

applied to a phosphocellulose column (about 1 mg of protein per ml bed

volume). The phosphocellulose column was washed with two volumes of buffer

A and the proteins were then eluted from the column using a linear gradient

(10 column volumes) of 0-600 mM NaCI in buffer A. Fractions equivalent to

5044

Downloaded from https://academic.oup.com/nar/article-abstract/8/21/5043/1063672by gueston 09 April 2018

Nucleic Acids Research

4% of the gradient volume were collected and assayed for endonuclease acti-

v i ty . The endonuclease activity which eluted at about 300 mM NaCI was

concentrated by dialysis against a 50% (v /v ) glycerol buffer which contained

20 mM Tris-HCI (pH 7.5), 300 mM NaCI, 0.1 mM EDTA and 10 mM p-mercap-

toethanol. After dialysis, the endonuclease was applied to a Sephadex G-75

column (1.68 x 68 cm) which had previously been equilibrated with buffer B

(20 mM Tris-HCI (pH 7.5), 300 mM NaCI, 0.1 mM EDTA and 10 mM p-mercap-

toethanol. The column was calibrated with ovalbumin, molecular weight

(m.w.) 45,000; chymotrypsinogen, m.w. 25,000; lysozyme, m.w. 14.000 and

RNase, m.w. 13,5000.

Preparation of 0X174 DNAs. The isolation of [3H]thymidine 0X174 RFI

and RFI I DNA was as earlier described (6). The viral form of 0X174 DNA

was purified by the method of Pagano and Hutchison (7). Unit length linear

0X174 DNA duplexes (RFIII) were generated by EndoR.Ava I (New England

Biolab) cleavage of the RFI form.

Endonuclease assay. The endonuclease assay was carried out in a reac-

tion mixture of 100-500 pi which contained 20 mM Tris-HCI (pH 7.5), 10 mM

p-mercaptoethanol, either 10 mM Mg or 10 mM Mn unless otherwise indi-

cated, and radioactive 0X174 DNA and enzyme in the amounts indicated in

the legends to figures. The reaction mixture was incubated at 37 for 15 min

unless otherwise stated. The endonucleolytic activity was monitored either by

the nitrocellulose filter technique as earlier described (6), or by alkaline

sucrose gradient centrifugation or gel electrophoresis of the DNA.

Alkaline sucrose gradient centrifuqation and gel electrophoresis.

The procedure used for alkaline sucrose gradient centrifugation of DNA was

as earlier described (4). Alkaline gel electrophoresis of DNA was performed

essentially as described by McDonell et a[. (8) using horizontal 1% agarose

gels.

Protein determination. Protein concentrations were determined using the

Bio-Rad Protein Assay Kit.

Polyacrylamide gel electrophoresis. Sodium dodecyl sulphate gel electro-

phoresis was carried out according to the slab gel technique described by

Laemmli and Favre (9).

RESULTS

Endonuclease activity associated with the FLV core. In order to show

that the endonuclease which is activated upon detergent disruption of FLV

(4) actually is located in the interior of the virus particle, a preparation of

5045

Downloaded from https://academic.oup.com/nar/article-abstract/8/21/5043/1063672by gueston 09 April 2018

Nucleic Acids Research

FLV was treated with proteinase K such that the viral envelope protein

(gp 71) and all nonviral proteins were digested (Fig. 1).

The viral cores obtained after the proteinase K digestion were then sedimen-

ted through a sucrose cushion in order to remove the proteinase K, and

subsequently disrupted by Triton x-100 and assayed for endonuclease acti-

vity (Fig. 2). The endonuclease activity proved to be completely resistant to

the proteinase K treatment of the virus particles (Fig. 2). These results

indicate that the endonuclease is primarily associated with the virus core.

Analogous results to those shown in Figure 2 were also obtained when Mg

was substituted for Mn in the reaction mixture.

Purification of virus associated endonuclease. The FLV associated endo-

nuclease was purified by DEAE- and phosphocellulose ion exchange chromato-

graphy, followed by gel filtration on a Sephadex G-75 column (see Materials

and Methods). The enzyme was eluted from the DEAE- and phosphocellulose

columns at concentrations of about 50 and 300 mM NaCI respectively (Fig. 3A

and B). Only one major peak of endonuclease activity was detected on each

column. When this activity was subsequently applied on a Sephadex G-75

column in the presence of 300 mM NaCI, one peak of endonuclease activity

was eluted at a molecular weight of approximately 40,000 (Fig. 3 C).

It is apparent from these results that the endonuclease activity does not

reside in the FLV RNA-directed DNA polymerase whose molecular weight is

R F'9- 1 - T n e effect o f proteinase K on° FLV. FLV pelleted from 1000 ml cell

culture was resuspended in 1 ml20 mM Tris-HCI, pH 7.5 (about 1 mg

_ . ^ ^ ^ & protein/ml) and divided into two equal9P #1'** ^^^P aliquots. One of the aliquots was then

incubated with 12 pg of proteinase Kfor 20 min. at 37 C. The virus partic-les in both aliquots were then pelleted

_ _ through a 20% sucrose cushion inP 3 U "•• ^ • • ^ ^ • • B 20 mM Tris-HCI (70,000 x g for 1 h)

and resuspended in 0.5 ml 20 mMTris-HCI (pH 7.5). The proteins in

— each of the two preparations werethen analyzed by polyacrylamide gelelectrophoresis as described in Mate-rials and Methods. (A) Proteins fromthe FLV preparation which had not

D 1 0 • • been treated with proteinase K and,- *• - (B) proteins from the FLV prepara-

tion which had been treated withproteinase K.

5046

Downloaded from https://academic.oup.com/nar/article-abstract/8/21/5043/1063672by gueston 09 April 2018

Nucleic Acids Research

Fig. 2. The effect on the FLV endo-nuclease activity of treating the viruswith proteinase K. The two virussamples (with and without proteinaseK treatment) prepared as described inthe legend to Figure 1 were bothdivided into two equal aliquotes(250 pi each). Triton X-100 wasadded to a final concentration of 0.2%to one of the aliquotes which hadbeen preincubated with proteinase Kand to one of the aliquotes which hadnot been treated with proteinase K.The endonuclease activity in each ofthe aliquotes was assayed in thepresence of 10 mM Mn and 0X174RFI [3H]DNA (0.02 pg/ assay), usingthe nitrocellulose filtration techniqueas described in Materials and Methods.Endonuclease activity in FLV prepa-ration which had been incubated withboth proteinase K and Triton X-100

( - • - ) , incubated without proteinase K and with Triton X-100 (-A-), incu-bated with proteinase K and without Triton X-100 (-0-) , and incubated withneither proteinase K nor Triton X-100 (-A-).

pi Extract

10 BO

Fraction number10 so

Fig. 3. Purification of the FLV endonuclease by chromatography on (A)DEAE-cellulose, (B) phosphocellulose and (C) Sephadex G-75 columns. Thepurification procedure was as described in Materials and Methods (startingwith about 5 mg of virus proteins). The endonuclease activity in 10 pi fromeach column fraction was assayed by the nitrocellulose filter technique usingsupercoiled 0X174 RFI [3H]DNA (0.02 pg/assay) as a substrate.

5047

Downloaded from https://academic.oup.com/nar/article-abstract/8/21/5043/1063672by gueston 09 April 2018

Nucleic Acids Research

about 84,000 (10).

When starting with a virus extract containing 5 mg of protein less than

1 jjg protein was obtained in the purified endonuclease fraction after gel

f i l tration. The purification of the enzyme resulted consequently in at least a

10a-fold increase in the specific activity. An exact determination of the

specific activity, however, was not obtained, since the low protein content

(< 0.1 (jg/ml) in the purified enzyme preparation made it difficult to measure

the protein concentration. The added amount of purified enzyme in the

following experiments will be quantitated in \i\ rather than \ig. The number

of \i\ refers in each case to the volume applied from a 10 ml preparation of

FLV endonuclease which had been purified from 5 mg virus protein.

Properties of virus associated endonuclease. The endonuclease was

characterized with respect to the effect various salt and divalent cation

concentrations had on the enzymatic activity. The presence of divalent

cations was required for the endonuclease to function, as 10 mM EDTA com-

pletely inhibited the activity.

The enzyme was active in the presence of either Mg or Mn and the

optimal concentration was the same for both cations (Fig. 4 A) . However,

Mn greatly enhanced the endonuclease activity (Fig. 4 A) . Salt concen-

trations above 50 mM inhibited the enzyme activity in the presence of either

M n ^ or Nig"*"1" (Fig. 4 B).

The FLV associated endonuclease converts 0X174 RFI DNA to the open

RFII form rather than to linear molecules as judged by gel electrophoresis of

nicked DNA (4). Hence, the endonuclease introduces single rather than

double-stranded breaks into DNA. Information about how extensively the

purified endonuclease nicked RFI DNA in the presence of either Mn or

Mg was obtained by analysing the sedimentation characteristics of the

nicked DNA under denaturing conditions (Fig. 5).

The sedimentation profiles in Figure 5A show that RFI DNA was nicked

extensively by the enzyme in the presence of Mn and the extent of nicking

correlated to the amount of enzyme which was added. In contrast to these

results, in the presence of Mg only a very limited number of nicks were

introduced into the supercoiled DNA by the purified enzyme and the number

of nicks did not increase significantly upon a 10-fold increase in the enzyme

to DNA ratio (Fig. 5B). Unit length linear single-stranded DNA was the

predominant DNA species in the alkaline gradient (Fig. 5B), indicating that

in the presence of Mg the main reaction product was a relaxed circular

duplex with one nick in either one or both strands. The data suggests that

5048

Downloaded from https://academic.oup.com/nar/article-abstract/8/21/5043/1063672by gueston 09 April 2018

Nucleic Acids Research

20 40mM MnCl2/MgCl2

eo ieomil KC1

Fig. 4. The effect of divalent cations and salt on the FLV associated endo-nuclease activity. (A.) Endpnuclease activity as a function of the concentra-tion of Mn (-•-) or Mg (-0-) in the reaction mixture. (B) Endonucleaseactivity as a function of the KCI concentration in the presence of 10 mMMn (-4-) or 10 mM Mg (-0-) . Each assay mix contained 0.02 pg 0X174RFI [3H]DNA and 20 pi or 80 ^ of the^purif ied FLV endonuclease whenassayed in the presence of Mn or Mg respectively. The endonucleaseactivity was assayed using the nitrocellulose fi lter technique as described inMaterials and Methods.

the endonuclease in the presence of Mg acts more readily on supercoiled

than relaxed DNA.

In order to determine the distribution on the 0X174 RFI DNA of the

sites which could be nicked by the purified FLV endonuclease in the pre-

sence of Mg the following experiment was performed. An excess of 0X174

RFI DNA was treated with the purified FLV endonuclease and the nicked

RFI I molecules were then separated from the RFI DNA by centrifugation in a

neutral sucrose gradient. The majority of the RFI I DNA molecules contained

only one single-stranded nick as revealed by the fact that an equal number

of circular and unit length linear single-stranded molecules were obtained

upon alkaline sucrose gradient centrifugation of an aliquot of the RFI I DNA

(data not shown). The remaining RFI I DNA was then divided into two equal

aliquots, one of which was treated with the EndoR. Aval endonuclease which

mtroduces one double-stranded break at a specific site on the 0X174 DNA

duplex. The DNA in both aliquots were then analyzed by gel electrophoresis

5049

Downloaded from https://academic.oup.com/nar/article-abstract/8/21/5043/1063672by gueston 09 April 2018

Nucleic Acids Research

30 •

Fraction Number

Fig. 5. Alkaline sucrose gradient centrifugation analysis of the FLV endo-nuclease activity on 0X174 RFI [3H] DNA. (A) The sedimentation profile ofDNA after incubating 0.2 H9 RFI DNA with 0 \i\ (.-%-), 100 \i\ (-0-) and150^1 (-4-) of the purified FLV endonuclease in the presence of 10 mMMn . (B) The sedimentation profiles of DNA after incubating 0.2 \ig RFIDNA with 0 \i\ ( - • - ) , 200 pi (-0-) and^500 pi (-A-) of purified FLV endo-nuclease in the presence of 10 mM Mg . The reaction was carried out at37 C for 15 mln. The alkaline sucrose gradient centrifugation was as des-cribed in Materials and Methods. The C and L denote circular and unitlength linear single-stranded 0X174 DNA respectively.

under denaturing conditions (Fig. 6).

The DNA which had been treated only with the FLV endonuclease

migrated as single-stranded circular and unit length linear molecules. About

50% of the DNA which had f irst been treated with the FLV endonuclease and

subsequently with the EndoR. Aval endonuclease migrated as unit length

single-stranded DNA. This was expected in view of the fact that prior to

cleavage by EndoR. Aval the majority of the RFI I molecules contained one

strand which was not nicked by the FLV endonuclease. The remaining DNA

migrated as expected if the regions in the DNA which were recognized by

the FLV associated endonuclease in the presence of Mg were evenly distr i -

buted on the circular 0X174 DNA molecule since no dominant peak other than

the unit length linear DNA was detected (Fig. 6). Both strands of the DNA

5050

Downloaded from https://academic.oup.com/nar/article-abstract/8/21/5043/1063672by gueston 09 April 2018

Nucleic Acids Research

Migration Com)

"Fig. 6. Alkaline gel electrophoresisanalysis of the distribution on 0X174RFI [?.H]DNA of the sites which maybe nicked by.the purified FLV endo-nuclease in the presence of Mg .0X174 RFI ?H-DNA (1 pg) was incu-bated 15 min with 700 \t\ of the FLVendonuclease in the presence of10 mM Mg . The nicked RFI I DNAwas then separated from the RFI DNAby sedimentation in a neutral 5-20%sucrose gradient (in 20 mM sodium-phosphate pH 7.0, 1 mM EDTA) usingan International SB-283 rotor andcentrifuging at 38,000 rpm for 8 h at15 C. After centrifugation the RFIIDNA was collected and concentratedby precipitation in ethanol. Onealiquot of this DNA was then analyzedby alkaline gel electrophoresis usinghorizontal 1% agarose gels ( - • - ) ,whereas another aliquote of the DNAwas treated with EndoR. Aval endo-nuclease prior to analysis by alkalinegel electrophoresis ( -0-) . The C andL denote circular and unit lengthlinear single-stranded 0X174 DNArespectively.

molecule could be nicked by the FLV endonuclease, since higher enzyme to

DNA ratios than used in this experiment resulted in the nicking of both

strands of the same DNA duplex (Fig. 5B).

Numerous of single-stranded nicks are introduced randomly into super-

coiled ColE1 DNA by the endonuclease associated with the p-subunit of the

AMV polymerase in the presence of Mn , whereas under similar conditions

the enzyme nicks linear DNA duplexes poorly (2). It thus seems that this

enzyme may have the property that it nicks circular duplexes more readily

than linear duplexes. A somewhat similar observation has been made with the

S1 endonuclease of Aspergillus oryzae, which cuts both supercoiled and

relaxed circular avian sarcoma virus DNA intermediates, but not the linear

viral DNA duplex (11). These observations suggest that some of the endo-

nuclease sensitive sites which are present in supercoiled DNA due to the

special structural configuration of this DNA may also to some extent be

present in open circular DNA duplexes, whereas they are not frequently

present in linear duplexes. With this in mind the ability of the FLV asso-

ciated endonuclease to nick linear as compared to circular DNA duplexes was

5051

Downloaded from https://academic.oup.com/nar/article-abstract/8/21/5043/1063672by gueston 09 April 2018

Nucleic Acids Research

studied. Prior to purification the FLV associated endonuclease activity be--

haved similarly to the AMV associated activity, since linear 0X174 duplexes

were not readily nicked when incubated with FLV lysates under conditions

which introduced many nicks into supercoiled circular 0X174 duplexes

(Fig. 7 A and B).

Moreover, even though approximately 0.5-fold fewer nicks were introduced

into the DNA when using open circular rather than supercoiled circular DNA

as the initial substrate (Fig. 7 A and C), the open circular DNA functioned

also much better as a substrate for the endonuclease in the FLV lysates than

the linear duplexes (Fig. 7 B and C). However, though the endonucleolytic

activity in the FLV lysates invariably nicked circular duplexes more readily

than linear DNA .and thus appears to resemble the endonucleolytic activity

associated with purified AMV polymerase, after purification the FLV asso-

Fig. 7. Alkaline sucrose gradient centrifugation analysis of the endonucleaseactivity in FLV extracts in the presence of Mn . The DNA sedimentationprofiles after incubating (A) supercoiled circular, (B) linear and (C) openrelaxed circular DNA duplexes with (-0-) and without (-•-) 25 ug of proteinfrom Triton X-100 (0.2%) disrupted FLV. The reaction mixture contained0.2 pg of 0X174 [?H]DNA and 10 mM Mn . The assay conditions and alkalinesucrose gradient centrifugation were as described in Materials and Methods.The C and L denote circular and unit length linear single-stranded 0X174DNA respectively.

5052

Downloaded from https://academic.oup.com/nar/article-abstract/8/21/5043/1063672by gueston 09 April 2018

Nucleic Acids Research

ciated endonucleolytic activity appeared to have lost this specificity since

linear and circular DNA duplexes were nicked about equally well when

treated with the purified enzyme (data not shown). At the present time it is

not clear whether this loss of specificity is due to a partial destruction of

the endonuclease during the purification or whether i t is due to the loss of a

factor which influences the specificity of the enzymatic activity. Prior to

purification the endonuclease nicked single-stranded circular 0X174 DNA

poorly, and this characteristic was, however, retained by the enzyme after

its purification (data not shown).

DISCUSSION

An endonuclease has been shown to be associated with FLV and the

enzyme has been purified more than 10?-fold. Its molecular weight was

determined by gel filtration to be about 40,000. The endonuclease activity

apparantly does not reside in the FLV RNA-directed DNA polymerase whose

molecular weight is about 84,000 (10).

An endonuclease activity resides in the p-subunit of the AMV RNA-

directed DNA polymerase (2,3). Some of the p-subunits undergo a cleavage

processing which results in the formation of the polymerase a-subunit and

the p32 protein. The p32 protein appears to have retained the p-subunit

endonuclease activity (1). The FLV associated endonuclease activity has

characteristics in common with the endonuclease activity associated with the

AMV p32 protein. Both of these enzymes introduce in most supercoiled DNAi, i.

duplexes only one single-stranded nick per strand when Mg is present,

whereas a great number of nicks are introduced into both strands in the

presence of Mn . Neither enzyme acts readily on single-stranded DNA.

The enzyme described in this study was the only major endonuclease

activity observed to be associated with FLV, which suggests that in contrast

to the AMV polymerase, the FLV polymerase does not contain an endo-

nuclease activity. This would be in agreement with the fact that in the

presence of Mn the purified FLV polymerase is unable to convert super-

coiled ColE- DNA into relaxed circular duplexes (10). However, it is possible

that the FLV endonuclease originates from the viral polymerase by a process

similar to the cleavage processing of the AMV p32 protein, but that the

cleavage occurs very early in FLV maturation. This would be in keeping

with the fact that the pol gene of murine leukemia virus has a much greater

coding capacity than what is required for the coding of the viral polymerase

(12).

5053

Downloaded from https://academic.oup.com/nar/article-abstract/8/21/5043/1063672by gueston 09 April 2018

Nucleic Acids Research

In the presence of Mg the purified FLV endonuclease appears to

introduce largely one nick per strand in supercoiled DNA duplexes. As de-

termined by analysis of the nicked DNA by alkaline gel electrophoresis, the

regions in the DNA where this nick was introduced were fairly evenly dis-

tributed on the circular DNA duplex. This is consistent with earlier results

obtained with an unpurified enzyme preparation (4). Our data are compatible

with the idea that the nick is introduced either completely aspecifically or at

one of a number of possible enzyme recognition sites which are fairly evenly

distributed on the supercoiled circular duplex. In contrast to the FLV endo-

nuclease, the AMV p32 protein has been reported to nick supercoiled DNA at

a very few preferred sites (1). This difference may be due to a difference

in the two viral enzymes, or it may be due to the fact that the supercoiled

DNA duplexes used as substrates in the two studies were different.

The fact that covalently closed circular virus DNA may be synthesized

[n vitro using detergent disrupted retroviruses (13) suggests that these

viruses contain all the enzymes necessary for the synthesis of the mature

viral DNA. A function for the virus associated endonuclease in this DNA

synthesis is not yet obvious. The most intriguing aspect of this endonuclease

is that in virus lysates in the presence of Mn** it seems to differentiate

between circular (RFI and II) and linear duplexes since the former molecule

is nicked more readily than the latter. A somewhat similar observation has

been made with endonuclease activity associated with AMV (2) and with the

S1 endonuclease (11), and it suggests that under certain conditions open

circular DNA duplexes may differ structurally from its linear counterpart in

a manner which enables some endonucleases to differentiate between them.

The nature of such a recognition site in the circular duplex is unknown. It

could not, however, purely be a base sequence which is being recognized,

since otherwise the endonuclease should also recognize the site in the linear

DNA molecule. The biological significance of this apparent specificity in case

of the FLV associated endonuclease, however, remains uncertain, since the

specificity was not retained upon purification of the endonuclease activity.

Nevertheless, the fact that the purified enzyme in the presence of Mg

appears to preferentially nick supercoiled DNA suggests that the circular

supercoiled proviral DNA may be the biological substrate for the endo-

nuclease and that it somehow participates in the replication of this DNA.

ACKNOWLEDGEMENT

We are grateful to Dr. T.S. Eikhom and Professor K. Kleppe for the

5054

Downloaded from https://academic.oup.com/nar/article-abstract/8/21/5043/1063672by gueston 09 April 2018

Nucleic Acids Research

interest they have showed in this work. The authors are Fellows of the

Norwegian Cancer Society (Landsforeningen mot Kreft) and this work was

supported by a grant from the same institution.

•Present address: Norwegian Food Research Institute P.O.Box 50, N-1432A"s-NLH, Norway

REFERENCES

1. Grandgenett, D.P., Vora, A .C. , and Schiff, R.D. (1978). Virology 89:119-132.

2. Golomb, M., and Grandgenett, D.P. (1979). J. Biol. Chem. 254: 1606-1613.

3. Samuel, K.P., Papas, T .S . , and Chirkjian, J.G. (1979). Proc. Nat.Acad. Sci. USA 79: 2659-2663.

4. Nissen-Meyer, J . , and Nes, I.F. (1980). Biochim. Biophys. Acta 609:148-157.

5. Seifert, E., Claviez, M., Frank, H., Hunsman, G., Schwarz, H., andSchafer, W. (1975). Z. Naturforschung 30c:698-700.

6. Nes, I .F., and Nissen-Meyer, J. (1978~7T~ Biochim. Biophys. Acta 520:111-121.

7. Pagano, J .S. , and Hutchison I I I , C.A. (1972). Methods of Virology 5:79-123.

8. McDonell, M.W., Simon, M.N., and Studier, F.W. (1977). J. Molec.Biol. 110: 119-146.

9. LaemmTTT U.K., and Favre, M. (1973). J. Molec. Biol. 80: 575-599.10. Moelling, K. (1974). Virology 62: 46-59.11. Hsu, T.W., Guntaka, R.V., and Taylor, J.M. (1978). J. Virol. 28:

1015-1017.12. Kopchick, J . J . , Karshin, W.L., and Arlinghaus, R.B. (1979). J. Virol.

30: 610-623.13. CTayman, C.H. , Mosharr, E.T., Anderson, D.L., and Faras, A.J .

(1979). Science 206: 582-584.

5055

Downloaded from https://academic.oup.com/nar/article-abstract/8/21/5043/1063672by gueston 09 April 2018

Nucleic Acids Research

5056

Downloaded from https://academic.oup.com/nar/article-abstract/8/21/5043/1063672by gueston 09 April 2018